DeepSeek Liang Wenfeng talks about open source: Only make 6 times the profit without working overtime. If you want to make 100 times, you will be constrained by open source. Recently, the internal conversation of DeepSeek investor exchange meeting leaked. The founder Liang Wenfeng gave a complete explanation of the underlying logic of the open source strategy.

DeepSeek Liang Wenfeng talks about open source: Only make 6 times the profit without working overtime, and only if you want to make 100 times will you be constrained by open source. Recently, the internal conversation of DeepSeek investor exchange meeting leaked, and the founder Liang Wenfeng gave a complete explanation of the underlying logic of the open source strategy. He made it clear that "it will definitely be open source, and the strongest model will also be open source." Because he saw no inevitable benefits of closed source, he explained the relationship between open source and commercial profits from a financial perspective for the first time. Open source will not affect 6 times profit, but it will restrict 100 times huge profits. Liang Wenfeng gave a clear financial logic: a ten-month payback corresponds to about six times profit. Under this premise, open source will have no impact on the business model. But if you want to make a hundred times the profit, open source will indeed have an impact, because third parties can independently deploy based on the open source model and achieve the same effect at twenty times the cost, thus lowering the pricing space. He positioned six times profit as DeepSeek's "restraint" strategy, believing that this was a way to make the company last longer, to have more opportunities in technological evolution, and to have a greater chance of becoming an AGI. Medium scale is a "sweet spot" that cannot be copied by others. Liang Wenfeng further dismantled the practical barriers to open source: Even if the model is open source and all principles are made public, the threshold for others to use it is still very high, and it is even more difficult to achieve the same low cost. He positioned the current scale of DeepSeek as a "sweet spot" - startups are too small and lack strength, large companies are too big and difficult to organize, and medium size is a unique advantage. He even bluntly said that the team "doesn't have to work overtime at all, because it's not that difficult." From the outside, it seems that they have chosen the hard mode to do the most difficult research, but in fact they have abandoned a lot in other places, and it is very powerful and easy. via AI News (author: AI Base)